The primary task accomplished during the current year of the project was the collection of data on children's friendships. Forty five classes in ten schools were selected for the sample. These classes ranged in racial composition from all white to all black with almost half of the classes having mixed racial composition. The classes also varied in instructional organization from highly structured classes with primarily whole class instruction to flexible classes with considerable individualized instruction. Background information was gathered on the children including age, sex, race or ethnicity, SES, absence behavior, etc. Two measures of academic achievement based on standardized tests were obtained from most of the classes. Questionnaire data on children's norms about friendship, educational aspirations, etc. were collected. Finally, longitudinal data on the children's friendship choices were gathered. Teacher questionnaires yielded information on grouping practices and classroom climate. A limited number of classroom observations were conducted to identify behavioral correlates of friendship choices. These data are presently being coded and keypunched in preparation for analysis next year.